Exploring AI in Education: Quickdraw with Google
Recently, I had the exciting opportunity to delve into the fascinating realm of AI in education through a course I’m taking with Ed Technology Specialists. In this assignment, I engaged with the concept of data drawings and explored the capabilities of Quick Draw, an AI-powered drawing recognition system. In looking at the site I unexpectedly found a sense of unity in humanity. As I scrolled down the list of 117,000 images of postcards all drawn by humans a sense of togetherness unexpectedly emerged. The amount of people who have all drawn this same item in a similar way is cute. Imagine all the different languages we speak, yet we all draw postcards like little prehistoric icons. In this moment we are together creating. Unity in the most unexpected place. The aim of this activity was to analyze the commonalities in data drawings and gain insights into how Quick Draw correctly or incorrectly identifies the objects being drawn. In this blog post, I will share my findings and reflections on this enjoyable assignment.
I could imagine a class of kinders through 5th graders enjoying this activity. I had never used Quick Draw before and found it to be easy to play. It asks you to draw everyday items in 20 seconds and the AI guesses what you drew. Below is an example of one of my drawings. The game reminds me of pictionary!
Analyzing the Data Drawings: I observed three commonalities in the data drawings for the picture frame- rectangle, lines, and box shape. This exercise allowed me to identify recurring characteristics and patterns that emerged across multiple drawings. By recognizing these commonalities, I gained a deeper understanding of the key features that contribute to Quick Draw’s recognition accuracy. It became evident that Quick Draw’s algorithm is designed to identify specific shapes, lines, and spatial elements to make accurate predictions. This understanding has practical implications for incorporating AI-powered image recognition systems into educational contexts.
Educators can leverage this knowledge to guide students in understanding the importance of specific features when creating visual representations. By emphasizing the inclusion of key characteristics like rectangles, lines, and box shapes, students can enhance their drawings’ recognizability to AI systems like Quick Draw. This exercise not only cultivates students’ artistic skills but also develops their understanding of the visual elements that contribute to effective communication and machine recognition.
Furthermore, recognizing the commonalities in the data drawings allows educators to design more targeted instruction and assessment strategies. By focusing on key features like rectangles, lines, and box shapes, educators can provide explicit guidance to students on how to create drawings that align with AI recognition algorithms which increases digital literacy. This approach can enhance students’ understanding of the relationship between their artistic choices and the successful interpretation by AI systems.
Quick Draw’s Guesses: Remarkably, Quick Draw correctly guessed all the drawings in my assignment. This speaks to the remarkable capabilities of AI in image recognition tasks, especially since I think my hot tub looks like a bowl of soup!
Factors Influencing Quick Draw’s Accuracy: Quick Draw’s ability to accurately identify drawings is influenced by several factors. One notable factor is the inherent variability in human drawings. Each individual may depict an object differently, leading to variations in shape, form, and detail. Consequently, Quick Draw may occasionally make incorrect associations due to these variations in human artistic expression. For instance, the rectangle shape of an envelope and a microwave can lead to confusion. In time-limited drawing exercises, participants often focus on capturing the most basic shapes, which can sometimes overlap with the characteristics of other objects.
Teaching Teachers about AI Overall, this assignment served as an engaging exploration of the potential and challenges of integrating AI and chatbots into educational settings. I think this activity could be beneficial to explaining the complex topic of neural networks to educational leaders and teachers. Even if one doesn’t have the basic understanding of artificial intelligence they can still have fun playing this game and gain some insight as to where this information comes from so that it becomes less scary. It’s a great jumping off point. Right now people fear artificial intelligence and feel like it’s creepy. A short engaging activity like Quick Draw can start to pull that fear away and offer a glimpse behind the curtain of artificial intelligence for teachers and school leaders.
As we continue to advance in the realm of educational technology, it is essential to embrace the opportunities offered by artificial intelligence while also being mindful of the limitations and potential biases that can arise. By critically examining and understanding these tools, we can harness their power to enhance teaching and learning experiences, promote creativity, and foster innovative educational practices.
Acknowledgment: Special thanks to ChatGPT, an AI language model developed by OpenAI, for its valuable assistance in creating this blog post.
More content at PlainEnglish.io.
Sign up for our free weekly newsletter. Follow us on Twitter, LinkedIn, YouTube, and Discord.
Exploring AI in Education: Quickdraw with Google was originally published in Artificial Intelligence in Plain English on Medium, where people are continuing the conversation by highlighting and responding to this story.
https://ai.plainenglish.io/exploring-ai-in-education-quickdraw-with-google-3a9fb0bab4b2?source=rss—-78d064101951—4
By: Jude Miqueli
Title: Exploring AI in Education: Quickdraw with Google
Sourced From: ai.plainenglish.io/exploring-ai-in-education-quickdraw-with-google-3a9fb0bab4b2?source=rss—-78d064101951—4
Published Date: Tue, 04 Jul 2023 03:24:50 GMT
